Learn RAG. The production version.
Moonlabs is the operator-led AI Academy in Derby. We run three live companies — Homemove, home.co.uk and homedata.co.uk — and we teach twelve students per cohort to ship a real AI product, sell it to a real customer, and raise on it. Three pillars: Coding, Commercials, Investment. Twelve weeks. £6,000.
Moonlabs is what we are. Two operators — James Freestone and Louis O’Connell-Bristow — who run Homemove, home.co.uk and homedata.co.uk, have raised £5m+ for their own companies, and have been AI-native since the week ChatGPT shipped.
The Academy is what we do. A twelve-week, in-person, twelve-student cohort in Derby. You build a real AI product. You sign a paid pilot. You write a deck and a financial model. You leave with a deployed system, a paying customer reference and a live investor pipeline. Coding, Commercials, Investment — the three pillars taught in equal weight every week.
Almost every RAG tutorial online ends at the same point: a notebook that retrieves three chunks and answers a question. That is the easy half. The hard half — chunking strategy at scale, hybrid retrieval, reranking, eval suites that catch regressions, observability that explains a wrong answer — is what the Academy spends twelve weeks on, on the same retrieval stack we run in our own production AI systems.
Coding · production retrieval
Postgres + pgvector by default; dedicated vector stores where they earn it. Chunking, embeddings, hybrid lexical + vector queries, reranking, observability. A deployed retrieval system by week twelve — with evals you wrote yourself.
Commercials · vertical-search as a productised offer
Every SMB with a knowledge base, a policy library, a CRM archive or a customer-support corpus is one good retrieval system away from a transformative AI surface — and most do not have one. Pricing a vertical-search retainer, the discovery call, a one-page pilot agreement, the first paying customer. A paid pilot by week six — retrieval is most useful when someone pays you to retrieve over their data.
Investment · raising on retrieval-AI
Glean ($4.6bn enterprise search), Vectara, Pinecone ($750m+), Weaviate, Chroma, Hebbia ($130m document retrieval), Sourcegraph, You.com, Perplexity ($14bn) — retrieval-AI is one of the largest funded categories of the cycle. Cap table, ten-slide deck, financial model. A live investor pipeline by demo day.
Common questions.
Do I need machine-learning experience?
No. Almost no useful RAG system in 2026 trains a model from scratch — the work is engineering on top of frontier models. That is what we teach.
What stack do you teach?
Postgres + pgvector is the default; dedicated vector stores appear where they earn their place. The blog has a useful piece on the trade-offs you should read before applying.
Will I learn agents and tool use as well as RAG?
Yes — agents and tool use are central to the broader Academy syllabus, and most production RAG systems are agentic at the edges. The agents course page is the sister page for the agent-shaped framing.
Will my project have real users by graduation?
Yes, if you do the commercial work. Week three is the “first five conversations” milestone; week six is the first paid pilot. By week twelve, paying users.
Can I bring my own dataset?
Yes — and most students do. The wedge work in week one shapes the system around your actual data and your actual users.
More Academy entry points.
The Academy is one course with many doors. Each of these pages is a different entry point into the same twelve weeks.
Build it. Sell it. Raise on it. In twelve weeks.
Tell us the data you would retrieve over and who would pay for it. James and Louis read every application personally and reply inside the week.
© 2026 Moonlabs Incubator. All rights reserved.